论文部分内容阅读
本文提出了以残差平方和SSE、均方差RMSE、相关系数平方R2、检验指标Q1、检验指标Q2等5个统计学检验指标来确定燃气轮机部件特性方程的最佳拟合次数。所得拟合方程既防止由于拟合次数太低所引起的结果粗糙,又避免了拟合次数过高,将测量数据中的噪声也纳入拟合模型。本文结果表明,综合运用拟合检验指标,可以避免由经验选择拟合次数的不科学问题,为特性方程最优拟合次数的选取提供了理论依据。
In this paper, the best fit times of gas turbine component characteristic equations are determined by using five statistical test indexes, such as square error of residual SSE, mean square error RMSE, square of correlation coefficient R2, test index Q1 and test index Q2. The resulting fitting equation not only prevents the result from being too rough due to too low fitting times, but also avoids the high fitting times, and adds the noise in the measured data into the fitting model. The results of this paper show that the comprehensive use of fitting test indicators can avoid the unscientific problem of choosing the fitting times by experience and provide a theoretical basis for selecting the optimal fitting times of the characteristic equations.